Hilbert Space to L2 Space: The Adjoint Operator
📂Hilbert SpaceHilbert Space to L2 Space: The Adjoint Operator
Theorem
Let {vk}k∈N be a sequence defined in a Hilbert space H. Let a bounded linear operator T:ℓ2→H be defined as follows.
T{ck}k∈N:=k=1∑∞ckvk
Then, the adjoint operator T∗:H→ℓ2 of T is represented as follows.
T∗v={⟨v,vk⟩H}k∈N
Moreover, for all v∈H,
k=1∑∞∣⟨v,vk⟩H∣2≤∥T∥2∥v∥H2
And likewise, for all v∈H,
TT∗v=k∈N∑⟨v,vk⟩Hvk,v∈H
Explanation
The space ℓ2 is of significant importance among lp spaces because it possesses an inner product, not just being an example of p=2 but becoming a particularly important case in the Banach space, which deals with functionals. Especially because it is isometrically isomorphic to separable Hilbert spaces, it’s very important.
Proof
From the definition of T,
⟨v,T{ck}k∈N⟩H=⟨v,k=1∑∞ckvk⟩H=k=1∑∞ck⟨v,vk⟩H
Since T∗:H→ℓ2 maps an element of v∈H to some sequence (T∗v)k∈ℓ2, it can be represented as follows.
T∗v={(T∗v)k}k∈N
Since the operator is linear and bounded,
∥T∗v∥2=(k=1∑∞∣(T∗v)k∣2)1/2≤∥T∗∥∥v∥H
This implies the following for all k∈N.
∥(T∗v)k∥2≤∥T∗∥∥v∥H
In other words, the mapping v↦(T∗v)k is bounded for all k∈N.
Riesz Representation Theorem
Let H be a Hilbert space. There exists a unique w∈H satisfying f(x)=⟨x,w⟩ and ∥f∥H∗=∥w∥H for the linear functional f∈H∗ and x∈H in H.
(T∗v)k=⟨v,wk⟩H
Therefore, according to the Riesz Representation Theorem, there must exist wk∈H satisfying the following. This means that T∗v can be represented like this for some {wk}k∈N⊂H.
T∗v={⟨v,wk⟩H}k∈N
Then, according to the definition of T∗,
k=1∑∞ck⟨v,vk⟩H====⟨v,T{ck}k∈N⟩H⟨T∗v,{ck}k∈N⟩ℓ2⟨{⟨v,wk⟩H}k∈N,{ck}k∈N⟩ℓ2k=1∑∞ck⟨v,wk⟩H
Summarizing,
k=1∑∞ck⟨v,vk⟩H=k=1∑∞ck⟨v,wk⟩H
Thus,
T∗v={⟨v,wk⟩H}k∈N={⟨v,vk⟩H}k∈N
Then, from the properties of the adjoint operator, since ∣T∣=∥T∗∥ and T∗ is bounded,
∥T∗v∥ℓ22≤∥T∗∥2∥v∥H2≤∥T∥2∥v∥H2
If rewritten in series form,
k=1∑∞∣⟨v,vk⟩H∣2≤∥T∥2∥v∥H2
Finally, from T{ck}k∈N:=∑k=1∞ckvk and T∗v={⟨v,vk⟩H}k∈N,
TT∗v=k∈N∑⟨v,vk⟩Hvk,v∈H
■